Commits · 3680d6f88b0a173297a9158240aeff0721a8c2f8 · change / sglang · GitLab

20 Aug, 2025 4 commits
- [feature] Rework Ascend NPU graph support (#9350) · 3680d6f8
  Even Zhou authored Aug 20, 2025
```
Co-authored-by: ronnie_zheng <zl19940307@163.com>
Co-authored-by: yezhifeng (D) <y00897525@china.huawei.com>
Co-authored-by: anon189Ty <Stari_Falcon@outlook.com>
Co-authored-by: Maksim <makcum888e@mail.ru>
Co-authored-by: ssshinigami <44640852+ssshinigami@users.noreply.github.com>
```
  3680d6f8
- Fix gpt-oss response api streaming issue (#9368) · f5154495
  Keyang Ru authored Aug 19, 2025
  
  f5154495
- Fix triton backend eagle illegal memory access (#9344) · e0ce171d
  Ke Bao authored Aug 20, 2025
  
  e0ce171d
- Fix mini lb timeout issue (#9369) · fe43e889
  fzyzcjy authored Aug 20, 2025
  
  fe43e889
19 Aug, 2025 5 commits
- Revert "[feature] Ascend NPU graph support (#8027)" (#9348) · f4fafacc
  Even Zhou authored Aug 20, 2025
  
  f4fafacc
- [Bugfix] fix kv buffer register & dp attention & deepepmoe (#9327) · 01d47a27
  chenxu140 authored Aug 20, 2025
  
  01d47a27
- enable marlin fp8 blockwise (#8990) · e483ab6d
  Enrique Shockwave authored Aug 19, 2025
  
  e483ab6d
- [Bug] Fix input arguments of flashinfer_trtllm_moe (#9317) · 3c2c9f6c
  Jiaqi Gu authored Aug 18, 2025
  
  3c2c9f6c
- support for interns1-mini (#9299) · a31ea448
  zxy authored Aug 19, 2025
  
  a31ea448
18 Aug, 2025 11 commits
- Tiny fix CI (#9306) · 5626e20b
  fzyzcjy authored Aug 19, 2025
  
  5626e20b
- [GLM4.1V and GLM4.5V] Add vision transformer num_dummy_head support: max tp=4 -> max tp=8 (#9059) · c2fbf60f
  Binyao Jiang authored Aug 18, 2025
  
  c2fbf60f
- [PD] Propagate internal server errors from aborted requests to clients instead... · 98b44e9e
  datdo-msft authored Aug 18, 2025
```
[PD] Propagate internal server errors from aborted requests to clients instead of blindly returning 200's (#8936)
```
  98b44e9e
- upgrade xgrammar 0.1.23 and openai-harmony 0.0.4 (#9284) · 6805f6da
  Swipe4057 authored Aug 19, 2025
  
  6805f6da
- [Docs] Correct and clarify notes in Engine docstring (#9313) · ca533580
  江家瑋 authored Aug 19, 2025
```
Signed-off-by: JiangJiaWei1103 <waynechuang97@gmail.com>
```
  ca533580
- [MISC] use dynamic choices for tool-call-parser argument (#9316) · 886454e8
  Keyang Ru authored Aug 18, 2025
  
  886454e8
- should return invalide request for empty prompt (#9315) · 0cf3fbeb
  gongwei-130 authored Aug 18, 2025
  
  0cf3fbeb
- Modelopt quant config adaptation (#8829) · 2256d62d
  Zhiyu authored Aug 18, 2025
  
  2256d62d
- Minor style fixes for sgl-kernel (#9289) · c480a3f6
  Lianmin Zheng authored Aug 18, 2025
  
  c480a3f6
- Further fix memory pool leak error (#9298) · 4c0bb411
  fzyzcjy authored Aug 18, 2025
  
  4c0bb411
- [Fix] Add undefined `update_tensor_inplace` function (#6307) · 716e6827
  b8zhong authored Aug 17, 2025
  
  716e6827
17 Aug, 2025 16 commits
- Set the default attention backend for GLM-4.5v to fa3 (#9245) · 84b30d9e
  zifeitong authored Aug 17, 2025
  
  84b30d9e
- [CPU] Fix TP padding issue on Phi-4 (#8289) · ebbb75e9
  blzheng authored Aug 18, 2025
  
  ebbb75e9
- Tiny make fp4 moe method parameters more static (#8520) · b498cd21
  fzyzcjy authored Aug 18, 2025
  
  b498cd21
- [fix]: fix cutlass moe ut and and Opt H20 cutlass groupGemm performance (#9272) · 0fc54b97
  kousakawang authored Aug 18, 2025
```
Co-authored-by: wanghanpei <wanghanpei@bytedance.com>
```
  0fc54b97
- Fix memory pool leak error (#9271) · b3c1f2e4
  fzyzcjy authored Aug 18, 2025
  
  b3c1f2e4
- Fix swa eagle verify accuracy for Triton backend (#9279) · be1a3cd9
  Ke Bao authored Aug 18, 2025
  
  be1a3cd9
- [chore] Clean up redundant lora_weight_names concept to simplify code (#9131) · 4b74c3fc
  Lifu Huang authored Aug 17, 2025
  
  4b74c3fc
- `from python.sglang.srt` -> `from sglang.srt` (#9268) · 3d77a318
  Netanel Haber authored Aug 17, 2025
  
  3d77a318
- model: support nvidia/Llama-3_3-Nemotron-Super-49B-v1 (#9067) · 845d12a9
  Netanel Haber authored Aug 17, 2025
```
Co-authored-by: Kyle Huang <kylhuang@nvidia.com>
```
  845d12a9
- Quick Fix GLM (#9264) · e47800e1
  Stefan He authored Aug 16, 2025
  
  e47800e1
- ci: simplify multi-modality tests by using mixins (#9006) · 1df84ff4
  Mick authored Aug 17, 2025
  
  1df84ff4
- Bug fix: use correct mm_items in embed_mm_inputs (#8893) · 66d6be08
  Binyao Jiang authored Aug 16, 2025
  
  66d6be08
- Combine fp4.py and mxfp4.py into one file and support dynamic mxfp4... · 1c1f8a11
  kk authored Aug 17, 2025
```
Combine fp4.py and mxfp4.py into one file and support dynamic mxfp4 quantization in mxfp4.py (#9049)
Co-authored-by: wunhuang <wunhuang@amd.com>
```
  1c1f8a11
- [PD] Support PD disaggregation with Prefill PP (#8846) · 384f8ab5
  Shangming Cai authored Aug 17, 2025
```
Signed-off-by: Shangming Cai <caishangming@linux.alibaba.com>
Signed-off-by: Shangming Cai <csmthu@gmail.com>
Co-authored-by: root <huzhiyuan@xiaohongshu.com>
Co-authored-by: Ying Sheng <sqy1415@gmail.com>
Co-authored-by: Francis <38564764+ssssnow@users.noreply.github.com>
Co-authored-by: zitto <zhjc1124@gmail.com>
```
  384f8ab5
- fix unexcepted answer in EAGLE mode (#9252) · 6a9d6ca3
  zyksir authored Aug 17, 2025
  
  6a9d6ca3
- [feature] Ascend NPU graph support (#8027) · 94371dbb
  VDV1985 authored Aug 17, 2025
```
Co-authored-by: ronnie_zheng <zl19940307@163.com>
Co-authored-by: yezhifeng (D) <y00897525@china.huawei.com>
Co-authored-by: anon189Ty <Stari_Falcon@outlook.com>
Co-authored-by: Maksim <makcum888e@mail.ru>
Co-authored-by: ssshinigami <44640852+ssshinigami@users.noreply.github.com>
```
  94371dbb
16 Aug, 2025 3 commits
- [CI] add deepseek w4a8 test on h20 ci (#7758) · 81da16f6
  Hank Han authored Aug 16, 2025
  
  81da16f6
- Fix DP load for embedding (#9165) · bc938ea1
  Brayden Zhong authored Aug 16, 2025
  
  bc938ea1
- Add fp4 quantize before all-gather for Flashinfer cutlass MoE DP (max throughput) (#7667) · eff4eb3f
  Trevor Morris authored Aug 15, 2025
  
  eff4eb3f
15 Aug, 2025 1 commit
- Fix nan value generated after custom all reduce (#8663) · 983aa496
  kk authored Aug 16, 2025
```
Co-authored-by: wunhuang <wunhuang@amd.com>
```
  983aa496